Reasons why JBOH sux: #158 No committed capital and #159 No order flow: "Yes, Knight/Trimark Profits by Trading 'Net Stocks, But the Question Is: Does It Profit Too Much? By Bill Alpert
(DT I notice you're not talking $20 any more, crying for just $10 now)
The nation's legions of online traders have made Amazon.com and Yahoo
into giddy stocks. Day traders have also made a great business for
Knight/Trimark Group, the market-maker executing more than 15% of trades
in Internet stocks.
Online traders may not know that their orders to brokers such as E*Trade
and Ameritrade get passed along to Knight/Trimark's computerized trade
stations in Jersey City, New Jersey. But in barely four years, Knight/Trimark
has become the world's largest market maker, handling the orders of heavily
advertised brokers like E*Trade and Discover Brokerage Direct. As those
brokers have grown, so has Knight/Trimark. It's a lot like the relationship
between online retailers and FedEx.
Many of the Internet's businesses have yet to show profits. Not so with
Knight. The firm has made lush profits, and its stock has shot from an
October '98 price of $4 to $64, and more recently it has been trading at
about $40. Two weeks ago, Knight/Trimark completed a secondary stock
offering of nine million shares at $35 each. In all, 6.6 million of those shares
were sold by brokers like E*Trade that originally established Knight/Trimark
to do their market-making.
Knight/Trimark's own shares have thrived on Nasdaq, says Chief Executive
Kenneth D. Pasternak, "because we have revenue growth, margin growth and
earnings growth and because we're associated with the Internet."
As the Wall Street "face" of seemingly mad online investors, Pasternak's firm
makes some more traditional brokers mad, too. Not only has Knight garnered
15% of all the trading volume on the Nasdaq Stock market, blowing past
market-makers like Charles Schwab's Mash Institutional or Herzog Heine
Geduld, but investment bankers blame Knight/Trimark for disorderly and
delayed opening trades in many recent new issues. With crammed queues of
online customers ready to pay any market price to get new "dot.com" stocks,
Knight/Trimark has snarled trading in ways that suggest to some brokers that
the firm is enriching itself at other people's expense, including its own
customers.
"That's pure poppycock," says Pasternak. The opening day crush for new
issues is an industry problem, he says, plaguing all traders. Knight and similar
market makers are just processing transactions for individual investors who
think Internet stocks are worth more than do Wall Street's pros, Pasternak
argues.
Knight/Trimark itself was founded by Wall Street pros like Pasternak, who
came aboard in 1995 from the Troster Singer unit of trading firm Spear Leeds
& Kellogg after concluding that old-line marketmakers like Troster were
falling behind the changing markets. Knight/Trimark set up shop as a
partnership of more than 25 brokers, including E*Trade, Morgan Stanley's
Discover unit and the Waterhouse Investors Services unit of
Toronto-Dominion Bank. Knight jump-started its order flow by paying
rebates to its broker-owners. Indeed, Knight/Trimark's trading volume
doubled in 1998 to 38 billion shares of stock on 41 million trades, up from 20
million trades in 1997.
Market makers like Knight/Trimark are
invisible to the investor who calls a
stockbroker or sends a message to
Ameritrade's computers. But such
stockbrokers usually turn around and
give the order to market makers.
Through a network of Nasdaq
computer screens, market makers post
the price at which they'll buy and sell a
stock, often using their own capital to
take the other side of an investor's
trade, holding the shares in inventory
while finding another investor. Although electronic communications networks
like Reuters Group's Instinet or the Brass Utility (bought last week by
Sungard Data Systems) now match about 20% of Nasdaq trades without a
market maker's mediation, most electronic communications networks don't
risk their own capital to facilitate trades. While an investor may languish
waiting to match a trade on Brass Utility, a market-maker like Knight might
take the trade immediately, using its own dough until the appropriate buyer or
seller comes along.
Market-making was an easier road to profits when there was a big spread
between the prices that market makers bid and asked for a stock. After
buying a share for $20, for example, the market maker often could sell it
immediately for 20 3/4, and pocket the spread. But a 1994 study by
economists William G. Christie and Paul H. Schultz said the absence of
odd-fractioned quotes by Nasdaq market makers suggested that spreads
were governed not by competition but by collusion. In 1996, the Justice
Department and the Securities and Exchange Commission obtained a
settlement from the National Association of Securities Dealers to end this
casino skimming racket. Subsequent SEC rules have cut in half most
market-maker spreads -- and profits. Lately, the New York Stock Exchange
and even the NASD itself have chatted with electronic communications
networks about teaming up to compete with traditional market makers.
Into this changing environment was born Knight/Trimark's predecessor,
Roundtable Partners, in 1994. Even after a July 1998 initial public offering of
stock and the recent secondary, Knight's brokerage firm founders still control
about a third of the firm. But given Knight/Trimark's steady profit growth,
public investors have been only too happy to get their hands on the firm's
shares.
Despite narrowing spreads on trades that shaved Knight/Trimark's average
net revenues per trade by 23.0% last year, to $8.51 from $11.05 in 1997,
explosive trading volume in Internet stocks has lifted the firm's operating
results. Total revenues surged to $356 million for 1998, up from just $227
million in 1997. Earnings in 1998 reached $51 million, or $1.40 a share, from
the year-earlier $1.17 cents. The firm's earnings weren't taxed until recently.
Had they been, 1998 earnings would have been $1.07 per share compared
with 1997 earnings of 67 cents.
Driving Knight's trading volume is the general growth in trading of Nasdaq
shares, which has tripled in the past five years. In particular, the firm has
benefited from the rise of online trading, which is climbing at a rate of 25% a
year. Happily for the firm, market research by the likes of Forrester Research
shows that the average online trader engages in five times the number of
trades as a traditional individual investor. Pasternak also credits his firm's
success to speedy execution of trades.
In all, Knight makes a market in 6,000 Nasdaq stocks, automating most
trades with computers. The Trimark piece of the business has become a
leading market maker in off-exchange trading of stocks listed on the New
York and American stock exchanges, a business previously led by firms like
Bernard L. Madoff and D.E. Shaw. Knight/Trimark plans to extend its
off-exchange trading to stocks listed in Europe, as well as stock options
traditionally traded on exchanges like the Chicago Board Options Exchange.
It intends to do this through an investment in a start-up called the International
Securities Exchange.
Those opportunities, and Knight/Trimark's powerful growth, inspire a "buy"
rating from Gary C. Craft, an analyst at BancBoston Robertson Stephens,
which led underwriters of Knight/Trimark's recent stock offering. Craft
predicts Knight/Trimark will earn $1.72 a share this year and $2.24 next. If
the shares trade at a multiple of 33 times next year's earnings, Craft thinks,
Knight/Trimark shares could hit $75 within the next six to 12 months. That's
quite a jump from the stock's recent price of about $40.
But with the sure profits of spread trading now bygone, market makers like
Knight/Trimark are increasingly trying to make a buck taking short-term
positions in their stocks, whether long or short. To manage the risk of such
proprietary trading, firms like Knight limit their exposure to any specific stock,
with automated alarms and the oversight of experienced trading managers.
Still, Knight/Trimark had between $73 million and $208 million riding on its
long and short positions, at the end of various months in 1998. The firm's long
and short positions offset each other in the aggregate, so net month-end
exposure ranged from $39 million long to $56 million short.
Yet Knight has cranked out remarkably consistent trading profits. Up to its
July 1998 IPO, the firm had suffered just three days of net trading losses. In
the first half of 1998, it averaged daily trading profits of over $1.1 million, with
no losing days. But as Internet stocks went nuts in the 1998 fourth quarter,
many investors feared that $20 moves in stocks like Yahoo might catch
Knight/Trimark at the wrong end of more proprietary trades.
Pasternak says the recent trading volumes and price volatility of Internet
stocks has made losing days a little more frequent, but allowed bigger gains
on the winning days. "We're right about 50%60% of the time," says the chief
executive, "and wrong about 40%-50%."
To the great joy of investors, Knight/Trimark had record profits in the quarter
ended in December of $18 million, or 33 cents a share, miles ahead of the 26
cents a share expected by Wall Street. Throughout the madness of
fourth-quarter Internet trading, Knight/Trimark handled 12 billion shares and
maintained its operating profit margin at 25%.
Knight's volume has continued to grow since December, says Pasternak, and
the firm handled more than 400,000 trades in just one day, January 11. But
also grown since December are complaints by underwriters that Knight
aggravates the way some new Internet stocks trade. Many dot.com IPOs like
Smith Gardner or WebTrends have spiked far above their opening prices in
their first day's trading, then closed precipitously below. Webtrends, a
Portland, Oregon, software firm, got $13 a share in its IPO on February 19,
but after opening at $35, its shares shot to $38, then closed at $27. No one's
happy with that kind of a debut.
"We're all dismayed at the ways these IPOs trade," said the chief financial
officer of a recent new issue. "Even though you close up 100% from the
pricing, you have a bad feeling about it."
Some underwriters say Knight has repeatedly snarled up trading in the first
minutes of new issues with high-priced opening bids that "cross" the market. If
an underwriter indicates that he has clients willing to sell shares on the opening
for $30, for example, Knight might post a bid to buy at $35, instead of simply
hitting the underwriter's offer price of $30. Under Nasdaq rules adopted in
January, such a "crossed market" prompts Nasdaq officials to delay the
opening of a new issue for 15 minutes, while market makers adjust their bids.
Knight's Pasternak says bankers single out his firm because it has the
strongest association with online investment demand. In hot Internet deals like
theglobe.com and Marketplace.com, online demand from investors far
outstripped the expectations of investment bankers. There's been a "major
disconnect," Pasternak says, between the price investment bankers think such
shares deserve and the price the public is willing to pay. Pasternak sat on an
industry committee that looked for solutions to such public buying panics, but
the committee adjourned without agreeing on a solution.
One investment banker, who doesn't blame Knight, says that market makers
are just conduits for amateur investors who bid unrealistic prices. The
recklessness of such investors, says this old Wall Street hand, is "like people
driving a car who have no automobile license."
But another investment banker says he can't think of another market maker
that crosses markets with high bids as frequently as Knight. He grumbles, "No
one is as blatant as they are." The rub, of course, is that wildly fluctuating
stock prices give market makers the opportunity to make more money.
Crazy openings of hot issues are to be expected, says the banker. More
annoying is when Knight's high bids delay the opening of a deal that was
harder to sell, because investors in such deals are more easily shaken.
Several traders noted that online clients of firms like Knight/Trimark often
place so-called "market orders" for hot IPOs, giving the market maker
freedom to fill the order at whatever price the market bears. Those traders
speculated that crossed markets might result from a market maker filling its
clients' orders at a high bid price with its own short sales, then profitably
covering the short sales at the lower prices offered by clients of the
underwriter. Without inspecting the market maker's books, however,
outsiders can't tell if it's making such proprietary profits at the expense of
clients.
Pasternak says such scenarios certainly don't reflect what happens at Knight.
The firm may take positions to accommodate queued up orders for new
issues, queues that sometimes stretch out to a half-million shares, but Knight's
proprietary positions never exceed a fraction of those queues. And profits for
a market maker aren't guaranteed by shorting into demand for every new
issue. "For every deal that went down, I could show you one that went up,"
says Pasternak. "If you think it's a lay-up, you're wrong."
The Knight chief executive says his firm has encouraged customers to use
"limit orders" to set price ceilings on their IPO orders. That was also the
advice of SEC Chairman Arthur Levitt in a special January 27 message to
online investors. Nevertheless, Pasternak says Knight's online customers
continue to send about half their orders in the form of market orders,
concerned perhaps, about missing Yahoo's next big move.
Conspiracy theories about Knight's trading and the size of its proprietary
positions has become part of Wall Street mythology, says Pasternak.
Knight/Trimark's computerized efficiency would seem to insure the firm a
place among the survivors on Wall Street's changing landscape. But if the
dot.com stocks get whacked for any prolonged period of time, life at Knight
won't be so cheery.http://interactive.wsj.com/articles/SB920678454763510000.htm |
